Up, down, or sideways: emplacement of magmatic Fe-Ni-Cu PGE sulfide melts in large igneous provinces

The preferential localization of Fe-Ni-Cu-PGE sulfides within the horizontal components of dike-sill-lava flow complexes in large igneous provinces (LIPS) indicates that they were fluid dynamic traps for sulfide melts. Many authors have interpreted them to have collected sulfide droplets transported upwards, often from deeper staging chambers. Although fine (<1-2 cm) dilute (<10%-15%) suspensions of dense (similar to 4-5 g/cm(3)) sulfide melt can be transported in ascending magmas, there are several problems with upward-transport models for almost all LIP-related deposits: (1) S isotopic data are consistent with nearby crustal sources, (2) xenoliths appear to be derived from nearby rather than deeper crustal sources, (3) lateral sheet flow or sill facies of major deposits contain few if any sulfides, (4) except where there is evidence for a local S source, sulfides or chalcophile element enrichments rarely if ever occur in the volcanic components even where there is mineralization in the subvolcanic plumbing system, and (5) some lavas are mildly to strongly depleted in PGE >>> Cu > Ni > Co, indicating that unerupted sulfides sequestered PGEs at depth. Two potential solutions to this paradox are that (i) natural systems contained surfactants that lowered sulfide-silicate interfacial tensions, permitting sulfide melts to coalesce and settle more easily than predicted from theoretical/ experimental studies of artificial/analog systems, and (or) (ii) sulfides existed not as uniformly dispersed droplets, as normally assumed, but as fluid-dynamically coherent pseudoslugs or pseudolayers that were large and dense enough that they could not be transported upwards. Regardless of the ultimate explanation, it seems likely that most high-grade Ni-Cu-PGE sulfide deposits in LIPs formed at or above the same stratigraphic levels as they are found.